The IC 5146 star forming complex and its surroundings with 2MASS, WISE and Spitzer

1.6 MB·

by  N. A. Nunes

The IC5146 star forming complex and its surroundings
with 2MASS, WISE and Spitzer
N.A.Nunes 1, C.Bonatto,1. and E.Bica,1
6 1Departamento de Astronomia, Universidade Federal do Rio Grande do Sul, Av. Bento
1 Gon¸calves, 9500
0 Porto Alegre 91501-970, RS, Brazil
2

n
a
J

6
Abstract
]
R
Throughout the last decade sensitive infrared observations obtained by the
S
SpitzerSpaceTelescopesignificantlyincreasedtheknownpopulationofYSOs
.
h
associated with nearby molecular clouds. With such a census recent studies
p
- have characterized pre-main sequence stars (PMS) and determined parame-
o
r tersfromdifferentwavelengths. GiventherestrictedSpitzercoverageofsome
t
s oftheseclouds,relativetotheirextendedregions,theseYSOpopulationsmay
a
[ represent a limited view of star formation in these regions. We are taking

advantage of mid-infrared observations from the NASA Wide Field Infrared
1
v Survey Explorer (WISE), which provides an all sky view and therefore full
9
coverage of the nearby clouds, to assess the degree to which their currently
7
2 knownYSOpopulationmayberepresentativeofamorecompletepopulation.
1
We extend the well established classification method of the Spitzer Legacy
0
. teams to archived WISE observations. We have adopted 2MASS photometry
1
as a “standard catalogue” for comparisons. Besides the massive embedded
0
6 cluster IC5146 we provide a multiband view of five new embedded clusters
1
in its surroundings that we discovered with WISE. In short, the analysis in-
:
v
volves the following for the presently studied cluster sample: (i) extraction
i
X
of 2MASS/WISE/Spitzer photometry in a wide circular region; (ii) field-star
r
a decontamination to enhance the intrinsic Colour-magnitude diagram (CMD)
morphology (essential for a proper derivation of reddening, age, and distance
from the Sun); and (iii) construction of Colour-magnitude filters, for more
contrasted stellar radial density profiles (RDPs).
Keywords:
Galaxy, Molecular Cloud, YSO, embedded clusters: general, photometry
Preprint submitted to ASSc January 7, 2016
1. Introduction
Star formation occurs inside Giant Molecular Clouds (GMCs) and sur-
veys of molecular gas in galaxies show that it is typically concentrated in
large complexes or spiral arm segments having sizes up to a kiloparsec and
masses up to 107 M (Solomon & Sanders,1985 and Elmegreen, 1993). These
(cid:12)
complexes may contain several giant molecular clouds (GMCs) with sizes up
to 100 parsecs (pc) and masses up to 106M . The GMCs in turn, con-
(cid:12)
tain much smaller scale structures that may be filamentary or clumpy on a
wide range of scales (Blitz, 1993; Blitz & Williams, 1999; Williams, Blitz
& McKee, 2000). The substructures found in GMCs range from massive
clumps with sizes of several parsecs and thousands of solar masses, which
may form entire clusters of stars, to small dense cloud cores with sizes of
the order of 0.1 pc and masses of the order of 1M (Mendoza, 1985; Cer-
(cid:12)
nicharo,1991; Larson, 1994; Williams, Blitz & McKee, 2000; Andre, Ward-
Thompson & Barsony, 2000; Visser, Richer & Chandler, 2002). The internal
structure of molecular clouds is partly hierarchical, consisting of smaller sub-
units within larger ones (Scalo,1990; Larson,1995; Simon, 1997; Stutzki et
al., 1998; Elmegreen, 2000). In particular, the irregular boundaries of molec-
ular clouds have fractal-like shapes resembling those of surfaces in turbulent
flows, and this might mean that the shapes of molecular clouds are created
by turbulence (Falgarone, Phillips & Walker, 1991; Falgarone, Puget & Per-
ault, 1992).
Most molecular clouds form stars, but very inefficiently, typically turning
only a few percent of their mass into stars before being dispersed. Despite
the strong dominance of gravity over thermal pressure, this low efficiency has
long been considered problematic, and implying that additional effects such
as magnetic fields, angular momentum conservation or turbulence support
these clouds in near-equilibrium against gravity and prevent a rapid collapse
(Heiles et al., 1993; McKee et al., 1993).
The lifetimes and evolution of molecular clouds is provided by the ages of
the associated newly formed stars and star clusters (Blaauw, 1991; Lar-
son,1994; Elmegreen, 2000; Andre, Ward-Thompson & Barsony, 2000; Hart-
mann, Ballesteros-Paredes & Bergin, 2001). Very few GMCs are known that
are not forming stars, and the most massive and dense ones as a rule contain
newly formed stars. They also cannot survive for long after beginning to
2
make stars, since the age span of the associated young stars and clusters is
never more than ∼ 10 Myr, about the dynamical or crossing time of a large
GMC. Stars and clusters older than 10 Myr do not appear to be associated
with molecular gas (Leisawitz, Bash & Thaddeus, 1989).
The youngest stars are associated with the denser parts of molecular clouds,
and especially with the densest cloud cores that appear to be the direct pro-
genitors of stars and stellar groupings (Mendoza,1985; Lada, Strom & Myers,
1993; Williams, Blitz & McKee, 2000; Andre,Ward-Thompson & Barsony,
2000).
After several decades of study, the physical conditions in nearby star-forming
molecular clouds are now fairly well understood, at least for the smaller
nearby dark clouds that form mostly low-mass stars. However in the past
40 years, infrared observations have revolutionized our understanding of star
formation (Gutermuth et al., 2009). The emission excess of young stars is
well above that expected from reddened stellar photospheres and originates
from the dusty circumstellar disks and envelopes surrounding young stars.
For these reasons, colour-colour diagrams and CMDs in the infrared (IR)
have proven to be excellent tools for identifying and classifying young stellar
objects (Megeath et al. 2004; Harvey et al. 2008; Gutermuth et al. 2009).
We intend to address the physical mechanisms responsible for the formation
of PMS stars in embedded star clusters which is crucial for understanding
their properties.
In Sect. 2 we discuss the IC5146 nebular complex to which a Streamer ap-
pearstoberelated,andcontainsasampleof5newlyfoundembeddedclusters
(ECs). In Sect. 3 we gather the data. In Sect. 4 we discuss the general prop-
erties of the objects. In Sect. 5 we discuss the results. Finally, in Sect. 6 we
provide the conclusion of this study.
2. Overview of the IC 5146 nebular complex
IC5146 is a reflection and emission nebula in the Cygnus constellation
centredataboutl = 94◦40(cid:48),b = −5◦50(cid:48). Totheeastofthecloudthereoccurs
aStreamer. TheStreamerregioncorrespondstoalongdarknebuladiscussed
by Wolf (1904). Hubble (1922) classified it in the optical as continuous,
whereas Minkowski (1947) listed the nebula to have Hα emissions. Both are
right, since IC5146 is a transition case between a reflection nebula and an
HII region. Many surveys at different wavelengths have been carried out in
this region and associated clouds (e.g. Kramer et al., 2003). Fig. 1 shows
3
NBB4
NBB3
NBB2
NBB5
NBB1
Figure 1: NOAO/AURA/NSF optical image from Kitt Peak telescope (Field is ∼ 2◦ ×
1.5◦)ofIC5146(left)andthepositionsofthefivenewECs(squares)alongtheStreamer.
North to the top and east to the right. Image credit: Adam Block.
an optical image of the IC5146 nebular and the Streamer (∼ 1.7◦ × 0.8◦)
extending to the east. In the present study, one of us (E.B.) discovered 5 new
ECs in the area of IC5146 using WISE (Fig. 2 and 3). The W2 images in
Fig. 2 enhance the stellar and protostellar components of the clusters, while
a W4 is sensitive to extended dust emission. Sources brighter in W4 than
in W2 are certainly due to the presence of dust envelopes or disks. NBB5
is projected close to IC5146 and all the clusters candidates apperar to be
immersed in cold molecular material. The central coordinates of these ECs
are given in Table 1, together with the estimated angular radii. Although
the cluster stellar concentrations may not be evident in Figs. 2 and 3, the
embeddedclusters, ingeneral, haveconspicuousRadialDensityDistributions
in Fig. 5.
TheIC5146regionhasanimportantpopulationofPMSstars. Embedded
sources, sometimes known as protostars are optically invisible young stars.
4
Table 1: The embedded cluster IC 5146 and the 5 newly found ECs along the Streamer
Name l b α δ Radius Number of stars
(deg) (deg) (h:m:s) (o :(cid:48):(cid:48)(cid:48)) ((cid:48)) (stars)
IC 5146 Cl 94.38 −5.49 21 : 53 : 26 47 : 16 : 11 7 162
NBB1† 93.49 −4.72 21 : 46 : 28 47 : 18 : 06 2 47
NBB2† 93.53 −4.26 21 : 44 : 52 47 : 40 : 32 5 24
NBB3† 93.60 −4.10 21 : 44 : 54 47 : 46 : 00 5 31
NBB4† 93.76 −4.64 21 : 47 : 22 47 : 32 : 10 8 45
NBB5†(cid:63) 94.24 −5.43 21 : 52 : 33 47 : 13 : 44 3 23
• (cid:63) close to IC5146
However, they are observed in the infrared and the majority is associated
with the cloud (see Sect. 3).We will explore IC5146 and the 5 ECs in the
nebular complex using Colour-colour diagrams and CMDs (Bonatto & Bica,
2007; Camargo, Bica & Bonatto, 2015). In this work, IC5146 refers either
to the nebula or the cluster.
3. The Data and Tools
The Infrared Array Camera (IRAC) was built at the NASA Goddard
Space Flight Center (Fazio et al. 2004). IRAC is the mid-infrared camera
on the Spitzer Space Telescope (Werner et al. 2004), with four arrays, or
channels, simultaneously taking data in two separate fields of view. The four
channels are referred to in this paper with their standard labels of 3.6,4.5,5.8
and 8.0µm for channels 1,2,3,4, respectively, as described in the IRAC doc-
umentation and by Fazio et al.(2004). It is a powerful survey instrument
because of its sensitivity, large FOV and four-colour imaging. IRAC on
Spitzer (e.g. Werner et al. 2004 and Fazio et al. 2004) has the potential
to extend our understanding of disk evolution and star formation by detect-
ing optically obscured, deeply embedded young stars and protostars. The
emission is detected from their disks, and, at earlier stages, from their in-
falling envelopes (Lada & Lada, 2003). The great advantage of IRAC over
ground-based telescopes is its sensitivity in the 3−8µm bands that contain
relatively little contribution from stellar photospheres as compared to disks
and envelopes. It is important to understand this colour space and use it to
identify young stars of various evolutionary classes (Allen et al. 2004). As an
5
NBB 1 NBB 3
NBB 2
NBB 4 NBB 5
Figure 2: WISE extractions in the W2 band (10(cid:48)×10(cid:48)) of the five newly found embedded
clusters. Upper panels from left: NBB 1, NBB 2 and NBB 3. Lower panels from left:
NBB 4 and NBB 5.
6
NBB 1 NBB 3
NBB 2
NBB 4 NBB 5
Figure 3: W4 band extraction (10(cid:48) ×10(cid:48)) of the five newly found embedded clusters
showing dust effects. Upper panels from left: NBB 1, NBB 2 and NBB 3. Lower panels
from left: NBB 4 and NBB 5.
7
embedded cluster IC5146 still remains partly embedded in the gas and dust.
The present discovery of 5 ECs in the area (Sect.2) further demands more
detailed studies around IC5146. We work with the 2MASS (J,H,K ) and
s
Spitzer (Ir1, Ir2, Ir3 and Ir4) photometry, which provide the spatial and pho-
tometric uniformity required for wide extractions. We employ the 2MASS
(J,H,K ) and Spitzer (Ir1, Ir2, Ir3 and Ir4) photometries, which provide
s
considerably higher resolution than WISE. The sources in the 2MASS data
are also cross-identified with a 0.5(cid:48)(cid:48) search radius. Only sources detected in
all four bands (W1, W2, W3 σ ≤ 0.1 and W4 σ ≤ 0.5) were initially consid-
ered. However, we realized that the W4 sources with profile magnitude > 8
should be taked with caution (Jarrett et al. 2011 and Wright et al. 2010).
We therefore limited our search for infrared excess to the range 3 − 12µm.
We did not use the WISE catalog flags to further filter our candidate list
because we found them in general unreliable for our purposes. In particular,
sources were often not flagged as extended or questionable, while no point
source actually existed. The source classification with WISE data indicated
source confusion in the central region and saturation effects, preventing iden-
tification of YSO candidates. We conclude that WISE may not be suitable
to classify YSOs in regions more distant than the IC5146 complex. In prin-
ciple, a random list of coordinates that falls inside a relatively smooth region
of nebulosity might be 100% classified as reliable. Counterparts in Spitzer
would support that. No attempt was made to compare measured flux levels
among WISE, 2MASS and Spitzer detections. This would require a surface
brightness measurement that WISE is not calibrated to produce. As a con-
sequence, we adopted 2MASS as standard catalog because the isochrones are
well determinated (Bressan et al.2012).
Within this perspective, our group has been developing analytical tools
for 2MASS photometry (Skrutskie et al., 1997) that allow us to statisti-
cally disentangle cluster evolutionary sequences from field stars in CMDs for
IC5146 and the Streamer (Bonatto & Bica, 2007). The decontamination
procedure is illustrated in Fig. 4 for IC5146. Decontaminated CMDs like
in Figs. 6 for IC5146 were used to investigate the nature of star cluster
candidates and derive their astrophysical parameters (Bonatto & Bica, 2006;
Bonatto et al. 2008). We apply the statistical algorithm described in Bon-
atto & Bica (2007b) to quantify the field-star contamination in the CMDs.
The algorithm uses relative star-count desity together and colour/magnitude
similarity between cluster and comparison field are taken simultaneously. It
measures the relative number densities of probable field and cluster stars in
8
cubic CMD cells whose axes correspond to the J magnitude and the (J−H)
and (J−K ) colours. The algorithm: (i) divides the full range of magnitude
s
and colours covered by the CMD into a 3D grid, (ii) calculates the expected
number density of field stars in each cell based on the number of comparison
field stars with similar magnitude and colours as those in the cell, and (iii)
subtracts the expected number of field stars from each cell. The algorithm
is responsive to local variations of field-star contamination (Bonatto & Bica
2007b). The adopted cell dimensions are large enough to allow sufficient
star-count statistics in the cells and small enough to preserve the morphol-
ogy of the CMD evolutionary sequences. For a representative background
star-count statistics we use the ring located within R ≤ R ≤ R around
inf ext
the cluster centre as the comparison field,where R usually represents twice
inf
the RDP radius. As extensively discussed in Bonatto & Bica (2007b), dif-
ferential reddening between cluster and field stars might be critical for the
decontamination algorithm. Large gradients would require large cell sizes or,
inextremecases, precludeapplicationofthealgorithm. Basically, itwouldbe
required, cell size δJ = 1.0 and δ(J −H) = δ(J −K ) = 0.2 between cluster
s
and comparison field for the differential reddening to affect the subtraction
in a given cell (e.g. Bonatto & Bica, 2008b). As a rule, a star cluster struc-
ture is studied by means of the stellar radial density profile (RDP). Usually,
star clusters have RDPs following a power law profile (King, 1962; Wilson,
1975; Elson et al., 1987). However, ECs are often heavily embedded in their
embryonic molecular cloud, which may absorb the near background and part
of the cluster member stars. As a result, some ECs may present RDPs with
decreasing density towards the cluster centre or bumps and dips (Lada &
Lada, 2003; Camargo, Bonatto & Bica, 2011, 2012; Camargo, Bica & Bon-
atto, 2013). Some RDP irregularities appear to be intrinsic to the embedded
evolutionarystageandrelatedtoafractal-likestructure(Lada&Lada, 2003;
Camargo, Bonatto & Bica, 2011, 2012; Camargo, Bica & Bonatto, 2013).The
Radial Density Profiles and CMDs (Figs. 5 and 6) determined in this study
are consistent with previous work of Embedded Clusters (e.g. Bonatto &
Bica, 2009, Bonatto & Bica, 2010).
Decontaminationisaveryimportantstepintheidentificationandcharac-
terizationofstarclusters. Differentapproaches(Merceretal.,2005)arebased
essentially on two different premises. The first relies on spatial variations of
the star-count density, but does not take into account CMD evolutionary
sequences. Alternatively, stars of an assumed cluster CMD are subtracted
according to similarity of colour and magnitude with the stars of an equal-
9
area comparison field CMD (as illustrated for the IC5146 in Fig. 4). In
general, CMDs of ECs are dominated by PMS stars, so decontamination is
importantforavoidingconfusionwiththereddwarfsoftheGalacticfieldand
background reddened stars. The IC5146 cluster 2MASS Color-color diagram
(Fig. 7) and that of Streamer (Fig. 10) are similar, which may suggest a sim-
ilar absorption law due to similarity between the medium and the structure.
A summary of different decontamination approaches is provided in Bonatto
& Bica (2009, 2010).
Withthissetup, thesubtractionefficiency, i.e. theaccumulateddifference
between the expected number of field stars (which may be fractional) and the
number of stars effectively subtracted (integer), over all cells is higher than
90% in all cases. The 2MASS decontamination results are given in Fig.6. In
short, theclusterdecontaminationwascarriedoutwiththeusualbackground
circumnuclear ring geometry that best described it. (e.g. Bonatto & Bica,
2011). These tools provided consistent RDP with observed profile of the
sources and CMD results (Figs. 5 to 7).
The MS stars of IC 5146 in the 2MASS CMDs (Fig. 6) correspond to the
blue isochrone, and they appear to have comparable reddening as the field
blue envelope of stars. The redder PMS stars populate the redder parts of
the CMD, as expected. The matched 2MASS-Spitzer diagrams (Fig. 8) of IC
5146 indicate a mixture of these populations, and/or some splitting between
cluster and field sources. The 2MASS and Spitzer CMDs and colour-colour
diagrams of the Streamer (Fig.10) show similarities with thoseof IC 5146,
with some MS stars and numerous PMS stars. The 2MASS photometry su-
perimposed with the PADOVA of the PARSEC version (1−5Myr) (Bressan
et al. 2012), indicates that the MS stars are clearly described,indicating an
ageing cluster. The PMS shows a wider colour range than the isochrone
distribution in Fig. 6. This implies that part of the sources are extremely
absorbed or have excess emission owing to envelopes or disks. The derived
parameters from the superimposed isochrones are given in Table 2 for IC
5146 and the additional objects (Sect. 4 and 5). Table 3 compares distance
determinations in the literature for IC 5146 to that derived in the present
study. Our distance agrees well with the larger values, e.g. in Herbig &
Dahm (2002).
10